Are current DFT methods sufficiently reliable for real-world molecular systems?

Abstract

Whereas density functional theory (DFT) based computations are the main tools in contemporary computational chemistry, there is an issue of chemical accuracy and reliability which, one the one hand, is recognised but, on the other hand, is not always sufficiently dealt with when it comes to application. Examples are given where DFT apparently fails. It is suggested that an effort needs to be undertaken to thoroughly investigate the limitations of current DFT calculations.